US 7729093 B1
A method and apparatus for detecting the presence of magnetic beads is disclosed. By providing both a static magnetic field and a magnetic field that alternates in the MHz range, or beyond, the bead can be excited into FMR (ferromagnetic resonance). The appearance of the latter is then detected by a magneto-resistive type of sensor. This approach offers several advantages over prior art methods in which the magnetic moment of the bead is detected directly.
1. A method to detect the presence of a magnetic bead, comprising:
providing a detector that signals the presence of a magnetic field through a change in its electrical resistance, said detector having an upper surface that is parallel to a free layer of the detector;
causing said free layer to be magnetized in a bias direction;
attaching said magnetic bead to said upper surface at a binding site;
providing a static magnetic field, having a value, in a direction normal to said upper surface;
providing an alternating magnetic field having a frequency, a first amplitude, and a direction parallel to said upper surface; and
varying said frequency and said static field value until ferromagnetic resonance (FMR) occurs, whereby said magnetic bead emits a magnetic field that rotates at said FMR frequency at a second amplitude, thereby causing said detector signal to peak and to signal the presence of said magnetic bead.
2. The method of
3. The method of
4. The method of
5. The method of
6. The method of
7. The method of
8. The method of
9. The method of
10. The method of
11. The method of
12. The method of
13. The method of
14. The method of
The invention relates to the general field of magnetic field measurement with particular reference to that of magnetic beads that have been selectively bound to biological structures and/or large molecules.
The invention discloses a methodology for detecting magnetic beads, magnetic particles and magnetic nano-particles by inducing ferromagnetic resonance (FMR) of their magnetic moment and by using a magneto-resistive (MR) sensor to detect the magnetic field produced by the FMR. Binding these magnetic beads or particles to biological or chemical molecules thus enables the presence of these molecules to be detected. In the form of a binding assay, where a matrix of MR sensors are patterned, this method can be used to identify the presence of a molecule of interest as well as to quantify its population. This method can address many other issues in the prior art that also utilize magnetic bead labeling and MR sensing, thus making this technology widely applicable.
Detection of biological or chemical molecules by using super-paramagnetic beads and particles as the labeling component and by using magneto-resistive sensors for detection of such labels is regarded as a promising technique to achieve accurate molecule counting with a resolution of several molecules or even a single molecule. It has the potential to enable fast, efficient, and economical biological applications, such as in-field virus and bacteria detection.
Binding assays to detect target molecules is a widely used technique in the biological, biochemical, and medical communities. The selective bindings commonly used include polynucleic acid bindings or hybridizations involving RNA and DNA, many types of ligand-to-receptor bindings, as well as antibody to antigen bindings. The target molecules in these bindings, for example, proteins or DNA, can also be a distinctive component or product of viruses, bacteria and cells, which may be the actual objects of interest for the detection.
In a binding assay, the binding molecules are attached to a solid substrate as “capture molecules”. When the assay is exposed to a liquid-form sample, where the target molecules attached to a physical label are contained, the binding molecules capture the target molecules with the selective bindings and immobilize the target molecules on the surface. This capture process is called “recognition”. The recognition events can be made to generate detectable signals from the attached labels and, consequently, the presence or absence of a target molecule can be detected.
In various prior art techniques utilizing the labeled binding process, the labels originally attached to the target molecule are also immobilized on the surface after the recognition process. The labels are either bound together with the target molecules on the surface (“sandwich” assay) or are by themselves (“competitive” assay). After the removal of non-bound labels, the bound labels can then be made to generate measurable signals.
When using magnetic beads or particles as the labeling component and a MR sensor as the detector, the MR devices are embedded below the binding surface and are usually covered by a protective layer. When the magnetic beads or particles are bound to the surface above a MR sensor, they can generate a magnetic field spontaneously or, in the case of super-paramagnetic beads or particles, if an applied magnetic field is present. This magnetic field from the beads or particles can be sensed by the MR device, which can then provide a voltage signal.
The magnetic labels described in published studies or as patents are usually super-paramagnetic beads—nano-particles (or larger beads that comprise nano-particles suspended in a non-magnertic matrix) that have no magnetization at room temperature without an externally applied magnetic field because of the super-paramagnetic effect. Such labels are desired in biological applications because they do not aggregate (at zero field). The beads or particles described in the prior art usually range in size from tens of nanometers to several microns. When the labels are attached to a surface after the recognition process, either single or multiple labels are attached to each MR device. However, the sensing mechanism has generally been the same for all the previous designs. When the magnetic labels are attached to the MR sensor top surface, the field generated by the magnetic moment of the beads or particles will either act directly on the underlying MR sensor or it may cancel a portion of the applied magnetic field acting on the sensor. For sensors that have no attached labels, the magnetic field from the magnetic labels is not present.
The pinned layer does not change its magnetization direction under normal magnetic fields. In a conventional MR sensor, a bias DC field Hbias is usually applied in the Y direction by a pair of opposing hard magnets on the sides of the sensor, so that the free layer's magnetization will be in the Y axis direction when no magnetic field is applied. However, this free layer alignment to Y axis can also be achieved by making the sensor dimension in the Y axis longer than in X axis due to the shape anisotropy. Because of the shape anisotropy, the magnetization of the free layer 2, as represented by Mfree, can only rotate freely in the XY plane when a transverse field is applied along the X axis direction and it is very difficult to rotate outside the XY plane, i.e. towards the Z axis.
If a magnetic field is applied in the X axis direction, the free layer magnetization rotates away from the Y axis and the resistance of the entire MR junction will change according to
where R0 is the base resistance of the sensor, ΔR is the full range resistance change of the sensor and θ is the angle between the magnetization of the reference layer and the free layer. With a DC current applied to the device, where the current can either flow in the XY plane or perpendicularly through the device, the voltage across the device will change, because of the resistance change, to produce a measurable voltage signal.
In prior studies and patents, several detection schemes were used. One commonly used scheme is to apply a magnetic field in the transverse direction [4-10, 12-13], e.g. along the X axis as in
It is important to note that these prior art methods use applied magnetic fields oscillating at frequencies less than 100 kHz (often much less) whereas the present invention requires frequencies in the MHZ (and higher) region.
A reference sensor to which beads will not attach to at any time is normally used for comparison of this voltage difference. During the detection, the applied field can also be a modulated by an AC field that will induce a same frequency AC voltage across the sensor. By utilizing a lock-in technique, the signal to noise ratio can be enhanced.
Another bead sensing scheme, also known as BARC [1-3,11], is to apply a DC field perpendicular to the film plane, i.e. along the Z axis direction in
A potential problem common to all the previously published or patented bead-MR sensor detection methods is that they are prone to fluctuations in the magnetic signal. Since all these detection methods are mainly practiced in the low frequency region from several Hertz to several kHertz, 1/f noise is very significant. Although the lock-in technique can successfully suppress the noise from the electrical sources by its narrow bandwidth, the noise from magnetic sources, for example Barkhausen noise, popcorn noise and telegraph noise cannot be prevented from affecting the locked-in signal. These magnetic noise sources are related to domain and local magnetization switching of MR sensors and are always most predominant in the low frequency region and usually show up as signal level random fluctuations. For the scheme that requires a reference sensor for detection, this combined noise effect from both the detection sensor and the reference sensor will at least double this parasitic fluctuation of the signal level.
Besides the fluctuations from the sensor, the bead itself will always have shape, size and magnetic content variations as well as binding site variations. These variations will also cause fluctuations of the amount of the bead magnetic field going into the sensor. With all the noise sources added together, these can be quite large and can cause significant signal fluctuations to inhibit practical binding assay applications that are based on the detection of the absolute field strength.
Another problem specifically for the field cancellation method is that a magnetic field needs to be applied in the sensing direction, i.e. X axis, to magnetize the magnetic beads and thus generate the cancellation field. However, this relatively large amplitude field will rotate the MR sensor free layer magnetization to the place where its sensitivity is not the highest. In other words, when the bead field is highest, the sensor sensitivity is lowest. By proper design of the MR sensor film structure and by using a vertical AC field to mimic the scissoring mode, second harmonic detection  can minimize the sensitivity loss. But the signal generated by the cancellation effect from the bead field is significantly decreased because it is only operating with a single bead field polarity and not a full reversal of the bead field direction in the transverse direction.
For the scissoring mode, where the bead is magnetized vertically, there is no sensitivity concern. However, this mode requires a relatively large sensor size. For a sub-micron or deep submicron size sensor, the exchange energy within the free layer will degrade the amount of rotation achievable for the two regions of the sensor rotating against each other. A low bias field, Hbias, or no bias field may be needed, which can easily lead to instability in a micron size sensor because of weak or no free layer domain control. In addition, for a GMR or TMR sensor with a single free layer, since this scheme only utilizes the sensor free layer magnetization rotation between 0 and 90 degrees, half of the sensor sensitivity region is not used.
For a multilayer GMR sensor, the rotation of magnetization can theoretically reach maximum or 0 to 180 degrees. A current-in-plane (CIP) multilayer GMR sensor usually has a lower dR/R, i.e. lower signal, than the state-of-the-art TMR or carefully designed spin valve GMR sensors. The current-perpendicular-to-plane (CPP) multilayer GMR sensors although possessing a much higher dR/R than the CIP ones, have also shown extraordinary magnetically related 1/f type noise in previous studies.
This noise can be more than 10 dB over the sensor's Johnson noise level in a micron size multilayer device and it will severely degrade the SNR of the sensor. To overcome the above problems, what is needed is, first, a scheme that can avoid having the bead magnetizing field affect the sensor free layer as well as utilizing the full reversal of the bead magnetization to gain maximum signal. Second, to avoid fluctuation of the bead magnetic field or the sensor resistance due to various magnetic sources, the method should not rely on detection of the absolute bead field magnitude. Third, signal detection at much higher frequencies than currently being explored (<100 kHz), for example beyond 1 MHz, is preferred in order to reduce low frequency noise effects.
A routine search of the prior art was performed with the following references of interest being found:
It has been an object of at least one embodiment of the present invention to provide a sensor for the detection of magnetic beads together with a method for utilizing said sensor.
Another object of at least one embodiment of the present invention has been that said sensor be relatively insensitive to the precise location of a bead relative to the sensor as well as to the exact size of the bead.
Still another object of at least one embodiment of the present invention has been that said sensor be relatively insensitive to the presence of other nearby beads.
A further object of at least one embodiment of the present invention has been that said method not require the measurement of the absolute value of a magnetic field generated by the bead.
These objects have been achieved by exciting FMR (ferromagnetic resonance) in the bead and then detecting the rotating magnetic field that the bead emits while in the resonance state. Our preferred means for detecting said rotating field has been a magneto-resistive detector such as a GMR (giant magnetic resistance) or TMJ (tunneling magnetic junction) device, though the method can be effectively used with any device capable of detecting magnetic fields as low as 1 Oe that are oscillating at frequencies in the MHz to GHz range. As a practical matter, the lowest field that can be used will be determined by the intrinsic noise discrimination and detection sensitivity of the sensor. Additionally, the detection sensitivity can usually be improved by narrowing the band width of the signal that is being detected.
The oscillating magnetic field used to excite the bead to FMR may be generated in a number of ways including external means and in-situ generation.
In the detection schemes to be described below, a uniform magnetization of the magnetic bead is assumed. The bead is only required to have a magnetic moment, either spontaneous or induced, when in the field.
Magnetic bead 1 lies above the top surface of the sensor. An externally applied static field HDC perpendicular to the sensor film plane orients the bead magnetization in the Z axis direction. Since the sensor only responds to X axis direction transverse fields, this DC field will not cause sensor resistance to change. Then, a low amplitude, high frequency, sinusoidal AC field hAC is applied in the Y axis direction to induce ferromagnetic resonance of the bead magnetization. This Y axis sinusoidal field does not itself cause the free layer magnetization to rotate as long as it is much smaller than Hbias.
Basis of the Invention:
As shown in
In first order calculations, the super-paramagnetic effect is not considered. For super-paramagnetic beads, a hAC field beyond 10 Oe may not be trivial for an HDC=50 Oe or 100 Oe. The AC field can also cause the bead moment to vary in magnitude instead of just direction.
Besides the comparison of the oscillation amplitude of bead moment at different frequencies or different DC fields, another way to detect the presence of a magnetic bead is to check the existence of an mx component at an AC field near or at a bead's resonant frequency at the given DC field. When there is no bead, no mx component is present. This method is an absolute signal level detection method so it is susceptible to bead field strength variations from various sources as in the prior art. However, it still has the advantage of detection at the bead FMR that produces the largest mx amplitude, i.e. highest detectable signal, using a relatively small excitation AC field.
The mx component of the bead magnetic moment will linearly influence the sensor's resistance change. Several previous studies [7,10] demonstrated a straightforward way to estimate the signal level from a GMR or TMR sensor with a given magnetic bead that has a magnetic moment in the X axis direction. The MR signal from the bead moment was reported to usually be a small portion of the entire dynamic range of the MR sensors, i.e. less than 1% to several percent of the dR/R. The signal of the bead FMR at the resonance frequency should be high enough to be picked up by available RF electronics with an applicable signal-to-noise-ratio (SNR). For example, in thermally excited FMR in GMR sensors, usually in the several GHz region and corresponding to less than 0.1% resistance fluctuation, has already been successfully measured with relatively simple RF circuitry . Therefore, FMR detection schemes proposed for monitoring the mx component of the bead magnetic moment should be achievable by measuring the MR sensor voltage output.
In summary, use of FMR to detect the presence of beads offers the following advantages:
1. The MR sensor responds to only the bead field during detection. The DC field for magnetizing super-paramagnetic beads and the AC excitation field can be applied in other than the sensing direction of the MR sensor. Thus there is no tradeoff needed between sensitivity and bead magnetic signal. In addition, it allows for adjustment of the sensor bias field towards lower values to enhance the sensor sensitivity to transverse fields.
This feature may be used as the basis for a method to determine the spatial location of a single bead which may, or may not, be part of an array of similar, or identical, beads. Se feature (g) below.
9. Since the resonance frequency is a function of the shape anisotropy of the magnetic beads and their damping constants, the method facilitates the simultaneous use of multiple beads whose resonance frequencies are not necessarily all the same, even in the same DC field. Examples include, but are not limited to, elongated magnetic beads and beads with a magnetic shell and a non-magnetic core. These structures will have different FMR frequencies from a spherical magnetic bead because of the shape anisotropy. It is also possible to vary the damping constant of the bead through control of its composition in order to shift the FMR frequency. In addition, magnetic beads formed from magnetic particles suspended in a nonmagnetic matrix will also exhibit different FMR resonance behavior, depending on the densities of the particles in the matrix.
All necessary layers, coatings and structures that enable the MR sensor to function in the relevant biological or chemical environments are assumed in the embodiments.
The magnetic beads used in the embodiments vary in shape, structure and composition as needed to obtain different bead FMR frequencies under the same DC field to enable bead labeling. The applied DC magnetic field in the following embodiments is also the field used to magnetize the beads should they be super-paramagnetic.
The MR sensor for FMR detection is not limited to GMR and TMR sensors, but rather any thin film sensing device that can show a measurable change in the presence of a magnetic field. The excited FMR of the bead moment produces an AC magnetic field in the MR sensor and subsequently an AC voltage signal across the sensor when a sensing current is applied. The detection of the presence of the bead can be achieved by the following methods (list not intended to be exhaustive):
(a) Sensor signal amplitude dependence on the AC field frequency when DC field is fixed. The amplitude is preferably measured at the same frequency as the AC field.
(b) Sensor signal amplitude dependence on the DC field when AC field frequency is fixed. The amplitude is preferably measured at the same frequency as the AC field.
(c) Sensor signal phase dependence on the AC field frequency when DC field is fixed. The phase should be measured at the same frequency as the AC field.
(d) Sensor signal phase dependence on the DC field when AC field frequency is fixed. The phase should be measured as close to the AC field frequency as possible.
(e) The existence of an AC signal from the sensor at the frequency of the driving AC field, preferably near or at bead's resonant frequency.
(f) Cancellation of the driving AC field on the sensor at a frequency preferably near or at bead's resonant frequency.
(g) Determining the precise location of a bead. In a conventional MRAM (magnetic random access memory), two non-parallel sets of conductive wires are used to form an array in which the intersection of any two wires (from opposing sets) is made to be unique by providing the same input to each wire, namely slightly more than half the magnetic field needed to trigger a single device. Consequently, it is only at the intersection of the two wires that the local field becomes strong enough to trigger a device.
In the present invention, the inputs provided to the two intersecting differ from one another. Referring now to
For methods (a) through (d), when a bead is present, the mentioned frequency or DC field dependence is observed. When there is no bead, no dependence is seen.
For method (e), the absolute value of the bead field is detected i.e. if there is no bead there is no field. This method is susceptible to bead field strength fluctuations caused by various sources. However, it still has the advantages that when detection is in the high frequency region, the MR sensor is at its maximum sensitivity and maximum bead magnetic field at a small FMR excitation field. Note that, for method (e) the AC field is not to be applied in the sensing direction of the MR sensor.
For method (f), the absolute combined total field of the applied AC field and the bead field is detected. When there is no bead, there is no bead field acting against the AC field effect on the sensor. This method is thus similar to (e) and enjoys the same advantages as (e) over the field cancellation methods of the prior art. In (f), however, the AC field is applied in the sensing direction of MR device.
For method (g), HDC is typically in the range of from about 1 and 1,000 Oe while hDC is typically in the range of from about 1 to 100 Oe. For example, for Hn=100 Oe and hn=20 Oe, FMR for beads on line 1 will occur at about 280 MHz while for a bead at the intersection of lines 92 and 93 it will occur at about 285.6 MHz.
Listed below are the preferred structural embodiments of this invention. Note that, although each of these depicts only a single bead over each sensor, it is clear that extension to multiple beads per sensor is readily implemented.
In addition, the phase of m, relative to hAC, will also show frequency dependence. With the MR sensor free layer resistance closely following the field from the mx, this phase dependence on the hAC frequency will also show up in the sensor's AC voltage signal and can be used as a detectable physical quantity as well.